Skip to contents

Cross-Validate Heterogeneous Treatment Effect Models

Source: R/cross_validation.R
cross_validate.Rd

Cross-validates a collection of heterogeneous treatment effect learners.

Usage

cross_validate(
  hte_learners,
  hte3_task,
  cv_metalearner = Lrnr_cv_selector$new(loss_squared_error),
  cv_control = NULL,
  ...
)

Arguments

hte_learners

A single Lrnr_hte learner, a list of Lrnr_hte learners, or a Stack of Lrnr_hte learners to cross-validate.

hte3_task

An hte3_Task object containing the data and necessary information for heterogeneous treatment effect estimation.

cv_metalearner

An optional metalearner (Lrnr_base object) used to combine the cross-validated learners. Default is Lrnr_cv_selector$new(loss_squared_error).

cv_control

A list of control parameters for cross-validation passed to Lrnr_sl. Default is NULL.

...

Additional arguments to pass to the loss function and other functions.

Value

A trained Lrnr_sl object containing the cross-validated ensemble of heterogeneous treatment effect learners.